HHS Public Access Author manuscript Author Manuscript
Mult Scler Relat Disord. Author manuscript; available in PMC 2017 September 22. Published in final edited form as: Mult Scler Relat Disord. 2016 September ; 9: 5–10. doi:10.1016/j.msard.2016.05.018.
Hospital admission rates for pediatric multiple sclerosis in the United States using the Pediatric Health Information System (PHIS)
Author Manuscript
Amy M. Lavery*, Brenda L. Banwell, Geraldine Liu, and Amy T. Waldman Children’s Hospital of Philadelphia, Colket Translational Research Building, 10012, 3501 Civic Center Blvd, Philadelphia, PA 19104, United States
Abstract Background—The onset of multiple sclerosis (MS) during childhood or adolescence is increasingly recognized in the United States. Administrative databases quantify healthcare utilization as measured by hospital admissions, providing insight into the impact of MS in the pediatric population. Objective—We examine the frequency of hospital admissions for pediatric MS in the US using the Pediatric Health Information System (PHIS) database.
Author Manuscript
Methods—Data was extracted from the PHIS database using the ICD-9 code for MS (340.00) and reviewed to verify case ascertainment. Mean, median, and range values were determined for the number of inpatient hospitalizations per patient, number of days in the hospital, and cost of each encounter. A trend analysis was performed to evaluate the annual frequency of MS-related admissions over the study period. Results—After case verification, the PHIS database extraction reported 2068 hospital inpatient encounters for 1422 unique pediatric MS patients between 2004 and 2013. The median number of hospitalizations per patient was 2 with a median hospital stay of 4 days. Admission rates for MS increased from 2.37 per 10,000 in 2004 to 4.13 per 10,000 in 2013. Conclusion—The number of admissions due to pediatric MS has increased since the start of the PHIS database collection, concurrent with increased disease awareness and the establishment of dedicated pediatric MS centers.
Author Manuscript
Keywords Multiple sclerosis; Pediatric; Pediatric Health Information System; Hospital admissions
1. Introduction Multiple sclerosis (MS) is a common neurological disease in adults, with an estimated prevalence of 90 per 100,000 people in the United States (National Multiple Sclerosis Society, 2016). Pediatric-onset MS occurs far less frequently, although creation of
*
Corresponding author. [email protected] (A.M. Lavery).
Lavery et al.
Page 2
Author Manuscript
international guidelines for MS diagnosis in children, (Krupp et al., 2007, 2013; Polman et al., 2011) increased access to MRI and identification of MRI features of MS in pediatric patients, (Verhey et al., 2011) and the establishment of formal pediatric MS clinics in the U.S. have led to increasingly frequent identification of MS in children and adolescents.
Author Manuscript
Pediatric MS is characterized by a relapsing-remitting disease course. Common relapse features include vision loss, sensory disturbances, weakness, balance difficulty, and gait impairment. These symptoms may result in an inpatient evaluation for diagnostic studies or hospitalization for intravenous (IV) corticosteroids or other therapy. While a survey of pediatric MS experts in the United States confirms that 97% of children with MS relapses are treated first-line with IV corticosteroids, these physicians also confirm that the decision to admit a child for IV corticosteroids is dependent upon multiple factors, such as relapse severity, timing of examination relative to the onset of relapse symptoms, and MRI findings indicative of disease activity (Waldman et al., 2011). Administrative databases quantify healthcare utilization and estimate the hospital-based burden of disease through quantification of admissions. The Pediatric Health Information System (PHIS) is an administrative database that contains inpatient, emergency department, ambulatory surgery, and observation encounter-level data from over 44 not-for-profit, tertiary care pediatric hospitals in the U.S. Several studies have used the PHIS database to determine hospitalization rates, cost burden to families, and further explore the impact of rare pediatric disorders across the U.S. (Wilson et al., 2015; Bourgeois et al., 2014; Akhavan et al., 2014; Olson et al., 2015).
Author Manuscript
We examined PHIS database-reported hospital admission rates and cost-burden for pediatric MS, and extrapolated the data to yield estimates of pediatric MS admissions and cost in the U.S. Using our institutional database to confirm the assumption that most pediatric MS patients are hospitalized at least once during their disease course, we then further extrapolated the national hospital admission rates to calculate disease prevalence estimates for pediatric MS in the U.S.
2. Methods
Author Manuscript
Data was extracted from the PHIS database for the years 2004 through the first 2 quarters of 2014 using the ICD-9 code for multiple sclerosis (ICD-9-340.00). The PHIS hospitals are affiliates of the Children’s Hospital Association (Overland Park, KS). Data quality and reliability are assured through a joint effort between the Children’s Hospital Association and participating hospitals. Portions of the data submission and data quality processes for the PHIS database are managed by Truven Health Analytics (Ann Arbor, MI). Data are deidentified at the time of data submission, and are subjected to a number of reliability and validity checks before being included in the database. Sites receive alerts if reviewed data is incomplete or appears to be of poor quality. For each PHIS entry with an ICD-9 code of 340.00, the following information was abstracted: demographics (sex, age, race/ethnicity), admitting diagnosis, procedure codes (including MRI and spinal tap), discharge diagnosis, payor source, and total cost of the
Mult Scler Relat Disord. Author manuscript; available in PMC 2017 September 22.
Lavery et al.
Page 3
Author Manuscript
admission. Only patients under the age of 19 years were included in the data abstraction. Case verification was determined by reviewing the admitting diagnosis, procedure codes, and discharge diagnosis for consistency with an MS indication. We excluded patients with codes for other demyelinating diseases such as acute disseminated encephalomyelitis or neuromyelitis optica and whose admission was for a cause other than an MS exacerbation (i.e. chronic tonsillitis). Records were sorted by discharge ID (a unique identifier) to determine discrete patients and the number of admissions per patient. For each admission, total cost reported in the PHIS database is adjusted for the Healthcare Finance Administration’s Center of Medicare and Medicaid Services wage/price index for the hospital location to help account for regional differences in payor mix and fee schedules. The total cost includes charges from inpatient stay, laboratory assessments, and imaging.
Author Manuscript
Mean, median, and range values were determined for the number of hospital inpatient admissions, number of days in the hospital, and cost of each encounter. Since encounters were only available for half of 2014, the trend analysis only included information from 2004 to 2013. Admission rates for MS were determined using the total amount of admissions reported within the PHIS system for the years 2004–2013 collectively and within each institution. A Mantel-Haensel analysis for trend was performed to compare yearly admission rates. All data analysis was performed using Stata 12.0 (College Station, TX).
Author Manuscript
The number of unique MS patients with inpatient hospitalizations was extrapolated to the U.S. population using census-level data provided by the CDC Wonder Surveillance Program (Center for Disease Control, 2015). These estimates rest on the assumption that each MS patient was admitted for inpatient care at least once during their disease course. To confirm this assumption, we evaluated hospitalization rates of patients treated at The Children’s Hospital of Philadelphia (CHOP) for MS. Medical records were reviewed from approximately the same time period as the PHIS database (2004–2013) to replicate a search based on coding data. Each patient chart was assessed to determine how many of the patients registered as outpatients in the Pediatric MS Center at CHOP, and living within the catchment area, were hospitalized at least once at CHOP during the same time period. This retrospective portion of the study was approved by the CHOP Institutional Review Board.
3. Results Information from 44 of the free-standing pediatric hospitals in the PHIS database reported at least one MS patient admission as described in Table 1.
Author Manuscript
Most of the hospitals contributed data prior to 2004 (86%) and 6 hospitals joined the PHIS network during the reporting period (2004–2014). The PHIS database extraction captured 1818 patients of whom 396 were excluded for having other diagnoses or non-MS related admissions. Of the 1422 remaining patients, 2068 hospital admissions for pediatric MS were reported between 2004 and the first quarter of 2014. Demographic information for all subjects is presented in Table 2. Approximately 13% of the patients were under the age of 11 years and the remainder were split evenly among the 11–15 year old age group (42%) and 16–18 year age group (45%).
Mult Scler Relat Disord. Author manuscript; available in PMC 2017 September 22.
Lavery et al.
Page 4
Author Manuscript
The female to male ratio was 2:1, but differed by age as shown in Table 3. Race was only reported for 29% of patients, of which 60.5% were white and 32.5% were black. Payor information was accurately recorded for 62% of patients, half of whom reported paying for care through Medicaid, Medicare or other government insurance. Hospital encounter information is presented in Table 4. Of the 1422 individual MS patients, the average number of admissions per patient was 6 (Median =4; Range 1–8). The median length of stay per admission was 4 days (Range 2–5). The median cost for an encounter was $24,672 (maximum cost = $323,1020.50). The annual rate of admissions for MS related events are presented in Fig. 1. Admission rates for MS increased from 3.47 per 10,000 in 2004 to 5.32 per 10,000 in 2013 (p
Mult Scler Relat Disord. Author manuscript; available in PMC 2017 September 22. Published in final edited form as: Mult Scler Relat Disord. 2016 September ; 9: 5–10. doi:10.1016/j.msard.2016.05.018.
Hospital admission rates for pediatric multiple sclerosis in the United States using the Pediatric Health Information System (PHIS)
Author Manuscript
Amy M. Lavery*, Brenda L. Banwell, Geraldine Liu, and Amy T. Waldman Children’s Hospital of Philadelphia, Colket Translational Research Building, 10012, 3501 Civic Center Blvd, Philadelphia, PA 19104, United States
Abstract Background—The onset of multiple sclerosis (MS) during childhood or adolescence is increasingly recognized in the United States. Administrative databases quantify healthcare utilization as measured by hospital admissions, providing insight into the impact of MS in the pediatric population. Objective—We examine the frequency of hospital admissions for pediatric MS in the US using the Pediatric Health Information System (PHIS) database.
Author Manuscript
Methods—Data was extracted from the PHIS database using the ICD-9 code for MS (340.00) and reviewed to verify case ascertainment. Mean, median, and range values were determined for the number of inpatient hospitalizations per patient, number of days in the hospital, and cost of each encounter. A trend analysis was performed to evaluate the annual frequency of MS-related admissions over the study period. Results—After case verification, the PHIS database extraction reported 2068 hospital inpatient encounters for 1422 unique pediatric MS patients between 2004 and 2013. The median number of hospitalizations per patient was 2 with a median hospital stay of 4 days. Admission rates for MS increased from 2.37 per 10,000 in 2004 to 4.13 per 10,000 in 2013. Conclusion—The number of admissions due to pediatric MS has increased since the start of the PHIS database collection, concurrent with increased disease awareness and the establishment of dedicated pediatric MS centers.
Author Manuscript
Keywords Multiple sclerosis; Pediatric; Pediatric Health Information System; Hospital admissions
1. Introduction Multiple sclerosis (MS) is a common neurological disease in adults, with an estimated prevalence of 90 per 100,000 people in the United States (National Multiple Sclerosis Society, 2016). Pediatric-onset MS occurs far less frequently, although creation of
*
Corresponding author. [email protected] (A.M. Lavery).
Lavery et al.
Page 2
Author Manuscript
international guidelines for MS diagnosis in children, (Krupp et al., 2007, 2013; Polman et al., 2011) increased access to MRI and identification of MRI features of MS in pediatric patients, (Verhey et al., 2011) and the establishment of formal pediatric MS clinics in the U.S. have led to increasingly frequent identification of MS in children and adolescents.
Author Manuscript
Pediatric MS is characterized by a relapsing-remitting disease course. Common relapse features include vision loss, sensory disturbances, weakness, balance difficulty, and gait impairment. These symptoms may result in an inpatient evaluation for diagnostic studies or hospitalization for intravenous (IV) corticosteroids or other therapy. While a survey of pediatric MS experts in the United States confirms that 97% of children with MS relapses are treated first-line with IV corticosteroids, these physicians also confirm that the decision to admit a child for IV corticosteroids is dependent upon multiple factors, such as relapse severity, timing of examination relative to the onset of relapse symptoms, and MRI findings indicative of disease activity (Waldman et al., 2011). Administrative databases quantify healthcare utilization and estimate the hospital-based burden of disease through quantification of admissions. The Pediatric Health Information System (PHIS) is an administrative database that contains inpatient, emergency department, ambulatory surgery, and observation encounter-level data from over 44 not-for-profit, tertiary care pediatric hospitals in the U.S. Several studies have used the PHIS database to determine hospitalization rates, cost burden to families, and further explore the impact of rare pediatric disorders across the U.S. (Wilson et al., 2015; Bourgeois et al., 2014; Akhavan et al., 2014; Olson et al., 2015).
Author Manuscript
We examined PHIS database-reported hospital admission rates and cost-burden for pediatric MS, and extrapolated the data to yield estimates of pediatric MS admissions and cost in the U.S. Using our institutional database to confirm the assumption that most pediatric MS patients are hospitalized at least once during their disease course, we then further extrapolated the national hospital admission rates to calculate disease prevalence estimates for pediatric MS in the U.S.
2. Methods
Author Manuscript
Data was extracted from the PHIS database for the years 2004 through the first 2 quarters of 2014 using the ICD-9 code for multiple sclerosis (ICD-9-340.00). The PHIS hospitals are affiliates of the Children’s Hospital Association (Overland Park, KS). Data quality and reliability are assured through a joint effort between the Children’s Hospital Association and participating hospitals. Portions of the data submission and data quality processes for the PHIS database are managed by Truven Health Analytics (Ann Arbor, MI). Data are deidentified at the time of data submission, and are subjected to a number of reliability and validity checks before being included in the database. Sites receive alerts if reviewed data is incomplete or appears to be of poor quality. For each PHIS entry with an ICD-9 code of 340.00, the following information was abstracted: demographics (sex, age, race/ethnicity), admitting diagnosis, procedure codes (including MRI and spinal tap), discharge diagnosis, payor source, and total cost of the
Mult Scler Relat Disord. Author manuscript; available in PMC 2017 September 22.
Lavery et al.
Page 3
Author Manuscript
admission. Only patients under the age of 19 years were included in the data abstraction. Case verification was determined by reviewing the admitting diagnosis, procedure codes, and discharge diagnosis for consistency with an MS indication. We excluded patients with codes for other demyelinating diseases such as acute disseminated encephalomyelitis or neuromyelitis optica and whose admission was for a cause other than an MS exacerbation (i.e. chronic tonsillitis). Records were sorted by discharge ID (a unique identifier) to determine discrete patients and the number of admissions per patient. For each admission, total cost reported in the PHIS database is adjusted for the Healthcare Finance Administration’s Center of Medicare and Medicaid Services wage/price index for the hospital location to help account for regional differences in payor mix and fee schedules. The total cost includes charges from inpatient stay, laboratory assessments, and imaging.
Author Manuscript
Mean, median, and range values were determined for the number of hospital inpatient admissions, number of days in the hospital, and cost of each encounter. Since encounters were only available for half of 2014, the trend analysis only included information from 2004 to 2013. Admission rates for MS were determined using the total amount of admissions reported within the PHIS system for the years 2004–2013 collectively and within each institution. A Mantel-Haensel analysis for trend was performed to compare yearly admission rates. All data analysis was performed using Stata 12.0 (College Station, TX).
Author Manuscript
The number of unique MS patients with inpatient hospitalizations was extrapolated to the U.S. population using census-level data provided by the CDC Wonder Surveillance Program (Center for Disease Control, 2015). These estimates rest on the assumption that each MS patient was admitted for inpatient care at least once during their disease course. To confirm this assumption, we evaluated hospitalization rates of patients treated at The Children’s Hospital of Philadelphia (CHOP) for MS. Medical records were reviewed from approximately the same time period as the PHIS database (2004–2013) to replicate a search based on coding data. Each patient chart was assessed to determine how many of the patients registered as outpatients in the Pediatric MS Center at CHOP, and living within the catchment area, were hospitalized at least once at CHOP during the same time period. This retrospective portion of the study was approved by the CHOP Institutional Review Board.
3. Results Information from 44 of the free-standing pediatric hospitals in the PHIS database reported at least one MS patient admission as described in Table 1.
Author Manuscript
Most of the hospitals contributed data prior to 2004 (86%) and 6 hospitals joined the PHIS network during the reporting period (2004–2014). The PHIS database extraction captured 1818 patients of whom 396 were excluded for having other diagnoses or non-MS related admissions. Of the 1422 remaining patients, 2068 hospital admissions for pediatric MS were reported between 2004 and the first quarter of 2014. Demographic information for all subjects is presented in Table 2. Approximately 13% of the patients were under the age of 11 years and the remainder were split evenly among the 11–15 year old age group (42%) and 16–18 year age group (45%).
Mult Scler Relat Disord. Author manuscript; available in PMC 2017 September 22.
Lavery et al.
Page 4
Author Manuscript
The female to male ratio was 2:1, but differed by age as shown in Table 3. Race was only reported for 29% of patients, of which 60.5% were white and 32.5% were black. Payor information was accurately recorded for 62% of patients, half of whom reported paying for care through Medicaid, Medicare or other government insurance. Hospital encounter information is presented in Table 4. Of the 1422 individual MS patients, the average number of admissions per patient was 6 (Median =4; Range 1–8). The median length of stay per admission was 4 days (Range 2–5). The median cost for an encounter was $24,672 (maximum cost = $323,1020.50). The annual rate of admissions for MS related events are presented in Fig. 1. Admission rates for MS increased from 3.47 per 10,000 in 2004 to 5.32 per 10,000 in 2013 (p
